New Track - Points vs Time?? Help!

[bushmg]

I will apologize in advance if this question has already been asked and answered 100's of times.

We have a small Pack (25 boys) and are set to run our PWD on 2/26. For the past 2 years, I have organized this event, and I will do it again this year.

Our old wooden track with "eyeball" finish line finally was replaced with a new aluminum 2-lane one (Best track). We got a Fast Track timer with serial interface and grandprix race manager software to talk to it.

In the past I used the SUPER Pope/Young PPN generator with pre-printed race charts (since I didn't know how many would show up beforehand). We ran the heats, tallied up the points, raced all the ties, and had our winners.

Now, we have another option - using the best average or cumulative times to determine the winners.

My question is this: how can I explain to a boy that wins all his heats that he's really in 3rd (or 4th) place because the other cars that he didn't race were faster than his car?

Has anyone else faced this problem? If you use times, the boys aren't really racing against the car in the other lane, are they? They're racing against all the other cars in the group.

Any suggestions or thoughts would be most appreciated!

Mike Bush

[Darin McGrew]

My question is this: how can I explain to a boy that wins all his heats that he's really in 3rd (or 4th) place because the other cars that he didn't race were faster than his car?

Before the first race, someone should explain to everyone present how the winners will be determined. For our PPN-like schedule, I would explain that everyone will race eight times, twice in each lane, and that everyone will race against as many different cars as possible, and that we'll determine the winners based on how many first-, second-, third-, and fourth-place finishes everyone has.

For a system with a timer, I would explain that everyone will race the same number of times in each lane, that we're going to time every car, and that we're going to use everyone's average time to determine the winners. Then, since you're changing the system from previous years, I would emphasize that the number of first-place finishes doesn't matter; all that matters is your average time.

You could also point out that everyone will get 4 races [adjust the numbers to fit your derby], against 3 opponents each race, so they'll be racing against only 12 other cars. But there are 19 cars total, so there are 6 cars that you'll never race against. You could win all your races, but if one of those 6 cars is faster, then that car will win based on time.

[Stan Pope]

My question is this: how can I explain to a boy that wins all his heats that he's really in 3rd (or 4th) place because the other cars that he didn't race were faster than his car?

Up front, I have a bias against "racing the clock" in Cub Scout racing. I won't let that admitted bias prevent me from trying to be objective.

A worse question to answer: "How can I explain to a boy who won all of his heats that he finished behind another car that he beat in one of those heats?"

Other possibilities, he wins 9 out of his 10 heats. Because of a relatively bad heat time in that lost heat, he finishes out of the trophies to cars that won fewer heats.

The cause is that a bad heat penalizes the racer very badly. It can be argued that a bad heat is a sign of a car that doesn't deserve to rate as high as a car that has no bad heats, even though its good heats were only fairly good. That is a matter of preference that your committee will have to thrash out.

While I prefer to limit the penalty for a bad heat by using a heat points method or an extensive elimination method, my main concern is that "racing the clock" can not be validated independently by the audience. When you ran Young&Pope (PPN) charts, the audience could tell if the cars were being rated accurately by the judges and if the points were being assigned correctly (if you kept score with transparencies on an ovefhead projector). I call that "having fairness and accuracy that is obvious to the casual ovserver." There is no way an audience member can validate the difference between two cars that receive times of 2.503 and 2.545 seconds unless the cars run in the same heat and they can do some quick computations in their heads.

If everyone trusts the guy running the clock and the guy running the starting gate to do so correctly every time, then they may accept racing the clock. A small mistake in gate operation can move cars down many places and be undetectable to most observers.

I guess it is time to get off my soapbox. Look at the comments in a related thread about using a time to break ties but main scoring is by points on a double round robin.

[Stan Pope]

Our old wooden track with "eyeball" finish line finally was replaced with a new aluminum 2-lane one (Best track). We got a Fast Track timer with serial interface and grandprix race manager software to talk to it.

In the past I used the SUPER Pope/Young PPN generator with pre-printed race charts (since I didn't know how many would show up beforehand). We ran the heats, tallied up the points, raced all the ties, and had our winners.

There is a variation of simple PPN charts that you may want to look at, especially if you have a large number of boys competing directly against each other. 20 boys on a 2 lane track makes for a long full round robin! (about 20*19/2 = 190 heats)


The alternative involves using a PPN chart of comfortable size to select finalists. Then run a 5 or 7 car PN chart among the finalists to determine trophy winners. If you have 2 place trophies to award, I'd suggest a 5 car PN (2 round) final (10 heat) chart. If you have 3 or 4 place trophies to award, I'd suggest a 7 car PN (3 round) final (21 heat) chart.

4 or 5 round PPN would be 80 or 100 heats where each of the 20 scouts race 8 or 10 times against 8 or 10 other scouts, which is a good number of heats per scout and total.

This plan is excellent for awarding the trophies to the objectively fastest cars in a way that everyone can tell is accurate. It is based on the fact that PPN charts perculate the best cars toward the top ... very high probability that the 3 fastest cars are among the 7 top scorers in a PPN chart.

I think that I recall that Randy's race management program supports this mode of operation.

If you want to evaluate the accuracy of the plan as you would apply it, get with me. I can give you some references to useful tools and examples of just such analysis.

[Derby Wizard]

If your district and council have races that the kids can advance to, you might consider following the same format used at those events. You can certainly do it differently. At this point we follow the same format used at district/council. Lowest average time - once in each lane at the pack level, the district and council runs each car twice in each lane since it has fewer lanes...rotating the cars through.

If we only had two lanes I believe we would do something different...but probably not a full round robin as we usually have >50 competing.

We would probably look into some type of PPN...

Derby Wizard

[bushmg]

WOW, what a fantastic response!!

Thank you all for your thoughtful and helpful replies.

Stan, I think your comment on audience validation really hit the nail on the head. In the past, I have been challenged by boys and parents who felt they had been treated unfairly. In each case, I showed them the scoring sheet and they were satisfied. The only time we had problems was when the finish was too close to call and we had to re-race or arbitrate to get a winner. Now, with the electronic finish line, we should have no further problems resolving close finishes. However, I'm not sure how I could answer the challenges if we used times, and I think there would be a lot more of them, too.

Stan, I would also like to take you up on your offer of some analysis tools for the 2-phase approach, since that is similar to what we do (i.e., the winners of each Den compete for the Pack Champ).

Again, thanks to everyone for the great ideas!

Mike

[Stan Pope]

If you take your plan through an exercise as described on my web page "Pinewood Derby Race Method Evaluation A CASE STUDY" at http://members.aol.com/standcmr/pwmevalc.html it will help you to understand and explain the strengths and limitations.

The page makes extensive use of some software constructed by Cory Young and posted on his pack's website. That software, in turn, derived, in part, from my earlier publiation, "Pinewood Derby Race Method Evaluation" at http://members.aol.com/standcmr/pwmeval.html. Cory added substantially with his statistical treatment of the information and his expression of lane differences.

[glaforge]

Stan and I will probably never agree about Total Time vs Place Points for determining the overall winners. I prefer the Total Time method.

I posted some results to this forum about a year ago. If I recall, both methods picked the same results. However, there were two cars that were so close in total time (1/1000th second) after 8 races one could have made an argument (and probably will) the results were too close to call or that the results should have been different. Last year’s results were pretty much the same so I didn’t post them.

One thing is for sure...If you choose place-scoring methods, somebody will come to you and tell you that their child should have placed higher because he had a lower total time. If you use total time, somebody will come to you and tell you their child should have placed higher because he won more races. Bet on it!

Now, as for why I like the total time method.....Plain and simple, I have had far fewer complaints. Period. I have not seen the problems that Stan eludes to...not that they can't happen...they just haven't. Another reason is it allows me to do some other creative things with scheduling without impacting the overall results. I just read a piece where somebody had software that was constantly figuring out the fastest cars and slowest cars and would create the heats matching them so that all the heats were close, giving the boys more of a thrill each race...can't do that with points but we sure can with total time! I can also schedule the boys to race only others at their own rank and still come up with an overall winner without having to run an additional race...To me it is important that all the boys run the same number of races (but that is a different thread).


And yes, racing the clock can be validated by the audience (and quite often is). We show the times each car raced in the heat just finished while we are setting up the next heat. Can they tell which car is faster in two different heats...yes they look at the times...can the tell if we simply put up what place they finished in...no, they both got the same score. They can also validate the results when they run head to head with that other car that is getting about the same times. It is simply the way you choose to look at it. People do seem to trust the start gate and electronic timers.

Whatever you choose, state how the places will be determined and let everybody know well in advance.

[Stan Pope]

And yes, racing the clock can be validated by the audience (and quite often is). We show the times each car raced in the heat just finished while we are setting up the next heat.

The "validation" that you describe is only "relative to the other times in the heat." Finish place tells you the same thing. It has no bearing on the overall validity of times in separate heats.

Here is what the audience can not validate:

If I run a car down the track and the timer tells you that the time is 2.555 sec. Then I run another car down the track and the timer tells you that the time is 2.554 sec. Who among your audience can validate either independent of your timer's report? (Hint: The answer is "None of them.")

Well, let's make it simpler ... the time for the car in heat 1 is 2.55 and for the car in heat 2 is 2.54. Who among your audience can validate the relative times, independent of your timer's report? (Hint: The answer is still "None of them.")

Okay, simpler still: the time for the car in heat 1 is 2.5 sec and for the car in heat 2 is 2.4 sec. Now, who among your audience can validate the relative times, independent of your timer's report? Some might be able to give very subjective comparisons of the car's speeds, but in the end they must rely on the report from the timer.

As starter, I can affect the times.

As "keeper of the timer", I can affect the times (in a lot of ways that I won't disclose.)

If the cars are staged by track staff, then that staff can affect the times.

As a participant, I can "inadvertently' bump the track and change its timing characteristics by amounts exceeding the timer's resolution. (I see that my car is running behind the front-runner. He is running straight and true down the track, I'm veering a bit left. Hmmm... the track is out of level! Next pass down, I'll bump the track a bit on its left side. Or, less conspicuously, I'll bump the starting line support a bit on its right side.)

Would I really do any of these things? No. But the plan leaves the opportunity open for "manipulations" that you may never catch.

The worst horror stories that I hear over the years come from folks who assumed that everyone involved followed the Scout Law. So, plan for everyone to follow the Scout Law, but arrange for the "bad apple" to have minimum opportunity to thrive.

[gpraceman]

Stan,

I understand your points, but I really think the scenario of someone intentionally bumping the track to give themselves an advantage is most unlikely. They could end up putting the track into better alignment for their opponents or could make it worse for their car.

To me many of the issues that could cause a variation in timing can be controlled. We have always done times scoring and after each race I take a look at the race data. I have yet to see any shifting of the times or any trends. I have gone as far as plotting the race data on a control chart to identify any trends in the data, including a shift and did not see any. We also end up with quite a few cars that perform with a very tight standard deviation over all their heats.

Personally, I think people should look at the issues involved and if they do not feel comfortable that their track setup and crew can perform consistently over the course of the race then they should use point scoring; otherwise I think scoring by times is fine.

[Stan Pope]

Stan,

I understand your points, but I really think the scenario of someone intentionally bumping the track to give themselves an advantage is most unlikely.

Personally, I think people should look at the issues involved

You did not comment on the "audience validation" aspect, which is the main point of the post. Does that mean you agree? Don't agree? or are reserving judgement?

I agree that each of these are "unlikely". I also agree that each organization should look at the issues and make up their own minds. In looking at the issues, they should not be impeded by "rose colored glasses." They should see all the issues and risks, then reach a reasoned conclusion consistent with their risk tolerance. Regarding the "risk tolerance" aspect, the organization needs to keep in mind that one failure offsets a lot of successes. What is the risk? What are the consequences if such an unlikely event occurs? Can your organization accept those consequences or should it seek to avoid them at reasonable cost? Should one of the risks apparently actualize, how should it be dealt with? Plan! Plan! Plan!

Related question: Suppose your post-race statistical analysis indicated a change in the track or some other non-random variances. From that analysis, you concluded that there is at least a 70% chance that you awarded the trophies to the wrong cars. What would you do? What could you do? The bell has been rung. Can you unring it?

[Darin McGrew]

One thing is for sure...If you choose place-scoring methods, somebody will come to you and tell you that their child should have placed higher because he had a lower total time.

We have never heard that complaint. Our finish gate reports only finish order, so no one can say anything about total time.

[gpraceman]

You did not comment on the "audience validation" aspect, which is the main point of the post. Does that mean you agree? Don't agree? or are reserving judgement?

I agree that the audience cannot validate the times unless they are doing timing on their own. However, I have never had an adult come up to me and dispute the results of a timed race.

I agree that each of these are "unlikely". I also agree that each organization should look at the issues and make up their own minds. In looking at the issues, they should not be impeded by "rose colored glasses." They should see all the issues and risks, then reach a reasoned conclusion consistent with their risk tolerance. Regarding the "risk tolerance" aspect, the organization needs to keep in mind that one failure offsets a lot of successes. What is the risk? What are the consequences if such an unlikely event occurs? Can your organization accept those consequences or should it seek to avoid them at reasonable cost? Should one of the risks apparently actualize, how should it be dealt with? Plan! Plan! Plan!

I agree with this a well. It is a judgment based on what risk an organization is willing to take and what is the probability of a problem occurring. To me scenarios like someone bumping the track to their advantage is a highly improbable event so the risk is very low. Inconsistancies in the timing system and/or the start gate release are much higher risk possibilities that should be looked at more closely.

Related question: Suppose your post-race statistical analysis indicated a change in the track or some other non-random variances. From that analysis, you concluded that there is at least a 70% chance that you awarded the trophies to the wrong cars. What would you do? What could you do? The bell has been rung. Can you unring it?

If I had noticed anything affecting the results, I would have implemented countermeasures. If those countermeasures did not alleviate the problem then I would have considered switching to points scoring. If trophies were ever awarded to the wrong kids, I would correct the matter by awarding additional trophies. I would never take a trophy away from a kid. Awarding trophies to the wrong racers is a scenario that I can easily see with points scoring as well.

From my experiences and observations I have not yet run into any justification to switch from times scoring. If I do, then I will reconsider my position.

[Stan Pope]

If trophies were ever awarded to the wrong kids, I would correct the matter by awarding additional trophies. I would never take a trophy away from a kid.

Excellent! The only answer in my opinion, too. I think you would have to remain publicly vague regarding the reasons, attributing to "possible timer malfunction" or somesuch and not even hinting at "manipulation."

Awarding trophies to the wrong racers is a scenario that I can easily see with points scoring as well.

But less likely if everything can be viewed and independently validated by the audience ... which was the point of my whole exercise.

[bushmg]

Again, thanks to all for your well-reasoned ideas!

Our race is Saturday 2/26, and I am planning on using the points scoring method for the following reasons.

First, this is the "shakedown cruise" for our new system. Although I have tested everything in my house, I want to get at least one derby done before I start trusting the equipment. Scoring by points allows me to have a back-up should there be an "equipment malfunction." We'll just keep racing using the old eyeball judging system. Randy's software is perfectly fine with switching to this method.

Second, it is more consistent with our past derbies. Everyone now understands the PPN/PN/CPN final standings system. I'll want to confer with the Pack Committee before switching horses.

Third, the racing IMHO is more intuitive and exciting when you are going "head to head" with a competitor in the other lane. With scoring by times, that isn't really the case. Each boy is really racing everyone else in the group, and across multiple heats. In fact, it might make more intuitive sense just to let each boy run his car alone through the lanes, then announce his total time (a la lane rotation).

Fourth, I really believe that it is important for the audience to be able to observe and validate the results of the scoring for each heat. This is easy to do with points, but harder with times.

Finally, we might try using times in a special race like an "open" or "Dad's" class after the Pack has finished. This would give us an opportunity to try it with lower consequences of failure.

I'll let everyone know how it went, and THANKS AGAIN for all the great discussion and input!

Mike Bush